WO2022209144A1 - 情報処理装置、情報処理方法及びプログラム - Google Patents
情報処理装置、情報処理方法及びプログラム Download PDFInfo
- Publication number
- WO2022209144A1 WO2022209144A1 PCT/JP2022/001145 JP2022001145W WO2022209144A1 WO 2022209144 A1 WO2022209144 A1 WO 2022209144A1 JP 2022001145 W JP2022001145 W JP 2022001145W WO 2022209144 A1 WO2022209144 A1 WO 2022209144A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- setting
- range
- data
- call
- conversation
- Prior art date
Links
- 230000010365 information processing Effects 0.000 title claims abstract description 16
- 238000003672 processing method Methods 0.000 title description 2
- 238000001514 detection method Methods 0.000 claims description 25
- 230000002123 temporal effect Effects 0.000 claims description 13
- 238000000034 method Methods 0.000 claims description 12
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 8
- 230000009471 action Effects 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 2
- 230000008685 targeting Effects 0.000 claims 1
- 230000004048 modification Effects 0.000 description 9
- 238000012986 modification Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 239000000203 mixture Substances 0.000 description 3
- 238000012790 confirmation Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000003058 natural language processing Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/04—Segmentation; Word boundary detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/42—Systems providing special services or facilities to subscribers
- H04M3/50—Centralised arrangements for answering calls; Centralised arrangements for recording messages for absent or busy subscribers ; Centralised arrangements for recording messages
- H04M3/51—Centralised call answering arrangements requiring operator intervention, e.g. call or contact centers for telemarketing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/211—Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
Definitions
- the present invention relates to an information processing device, an information processing method, and a program.
- a mechanism called time authentication has been conventionally known (for example, Non-Patent Document 1). By performing time authentication on electronic data, it becomes possible to prove that the electronic data existed at a certain time and that the electronic data has not been tampered with since that time.
- An embodiment of the present invention has been made in view of the above points, and aims to reduce the cost required for time authentication.
- an information processing apparatus includes a specification unit that specifies part or all of conversation data representing the conversation as a setting range for time authentication according to the content of the conversation of a plurality of people. and a setting unit that sets time authentication for the specified setting range.
- the cost required for time authentication can be reduced.
- the call data may include not only voice data in the call between the contact center operator and the customer, but also text data obtained by converting the voice data by voice recognition.
- the contact center is just an example, and in addition to the contact center, for example, the voice of the conversation that occurred during sales work and counter work for sales personnel such as products and services and counter personnel at stores The same can be applied to cases such as time authentication for conversation data including data and text data.
- time authentication for conversation data including data and text data.
- the operator of the contact center conducts business such as responding to inquiries by voice communication with customers, but the present invention is not limited to this. (including those that can be sent and received), video calls (including those that can share the display of electronic data), etc. (In other words, call data does not include voice data, only text data case) can be applied in the same way. More generally, it is possible to similarly apply the time authentication to conversation data relating to conversations of a plurality of persons.
- time stamping means that the time stamping authority of a trusted business operator (certified time stamping business operator) gives time information to the hash value (or message digest) of electronic data.
- a hash value to which time information is attached is called a time stamp token (or simply referred to as a time stamp). It proves that the electronic data has not been tampered with since then.
- performing time stamping on certain electronic data is also referred to as "setting time stamps on electronic data”.
- FIG. 1 is a diagram showing an example of the overall configuration of a contact center system 1 according to this embodiment.
- the contact center system 1 includes an authentication management device 10, an operator terminal 20, an administrator terminal 30, a PBX 40, a customer terminal 50, and a time stamping authority 60. included.
- the authentication management device 10, the operator terminal 20, the administrator terminal 30 and the PBX 40 are installed in the contact center environment E, which is the system environment of the contact center.
- the contact center environment E is not limited to the system environment in the same building, and may be, for example, system environments in a plurality of geographically separated buildings.
- the authentication management device 10 extracts a necessary range (hereinafter also referred to as a setting range) from at least one of voice data in a call between a customer and an operator and text data obtained by converting the voice by voice recognition, Set time authentication for this setting range. That is, the authentication management device 10 calculates the hash value of the data in the setting range, and requests the time stamp token of the hash value from the time stamping authority 60 via the communication network 80 including the Internet. And the authentication management apparatus 10 preserve
- a necessary range hereinafter also referred to as a setting range
- whether voice data, text data, or both are subject to time authentication, and which range is set as the setting range is determined according to the content of the call represented by the call data. More specifically, it is determined according to the call reason of the call represented by the call data and the terms uttered during the call.
- the call reason is the reason why the customer called the call center (the reason for the incoming call), and is the content of the customer's inquiry. For the sake of simplicity, the following description assumes that one call has only one call reason.
- the operator terminals 20 are various terminals such as PCs used by operators who respond to inquiries from customers, and function as IP (Internet Protocol) telephones.
- PCs used by operators who respond to inquiries from customers
- IP Internet Protocol
- the administrator terminal 30 is various terminals such as a PC (personal computer) used by an administrator who manages operators (such an administrator is also called a supervisor).
- PC personal computer
- the PBX 40 is a telephone exchange (IP-PBX) and is connected to a communication network 70 including a VoIP (Voice over Internet Protocol) network and a PSTN (Public Switched Telephone Network).
- IP-PBX telephone exchange
- VoIP Voice over Internet Protocol
- PSTN Public Switched Telephone Network
- the customer terminals 50 are various terminals such as smart phones, mobile phones, and landline phones used by customers.
- the time stamping authority 60 is a time stamping authority operated or managed by a time stamping business operator. In response to the request from the authentication management device 10, the time stamp authority 60 returns a time stamp token obtained by adding time information to the hash value included in this request.
- the overall configuration of the contact center system 1 shown in FIG. 1 is an example, and other configurations may be used.
- the authentication management device 10 is included in the contact center environment E (that is, the authentication management device 10 is an on-premise type), but all or part of the functions of the authentication management device 10 are It may be implemented by a cloud service or the like.
- the PBX 40 is an on-premise telephone exchange, but may be realized by cloud services.
- the operator terminal 20 functions as an IP telephone, for example, a telephone may be included in the contact center system 1 in addition to the operator terminal 20 .
- a device for example, a gateway device or the like
- that connects the contact center environment E and an external network may function as the authentication management device 10 .
- time authentication is set for a required range (set range) of required data (hereinafter also referred to as target data).
- target data is the data to be set for time authentication, out of the voice data and text data included in the call data.
- setting range is the range to be set for time authentication within the temporal range of the target data.
- FIG. 2 is a diagram for schematically explaining an example of time authentication settings.
- a certain important term W1 is uttered at time T1
- a certain important term W2 is uttered at time T2 in a call related to the call reason “product information”.
- voice data and text data are the target data
- the temporal range from time T 1 - ⁇ T 1 to time T 1 + ⁇ T 1 that is, the range of ⁇ T 1 before and after the utterance time of word W 1 ) is set.
- Time authentication is set as Also, for example, text data is the target data, and the time range from time T 2 - ⁇ T 2 to time T 2 + ⁇ T 2 (that is, the range of ⁇ T 2 before and after the utterance time of word W 2 ) is set as the set range. set. This is because the contents of the speech before and after an important term is spoken can be important matters, so it is appropriate to time stamp the contents of the speech.
- Important terms are important words and phrases related to the call reason. For example, when responding to an inquiry about a call reason "product information", the operator needs to fully explain to the customer the charge and warranty details of the product. Therefore, in this case, words such as “fee” and “guarantee” are important terms. Hereinafter, "important terms” may be simply referred to as “important terms.”
- the example shown in the right diagram of Fig. 2 is when a call is made regarding the call reason "Conclusion of contract".
- time authentication is set with voice data and text data as target data and the entire time range from the start to the end of the call as the set range. This is because, for example, in a call related to "Conclusion of a contract", all utterances can be important matters, so it is appropriate to perform time authentication for all text data and voice data from start to finish. is.
- one or both of voice data and text data are targeted for a certain temporal range (including all) according to call reasons and important terms spoken during a call. to set the time stamp.
- This makes it possible to set time stamps for appropriate ranges of data according to the call reason and the importance of important terms spoken during the call, e.g.
- the cost required for time authentication can be reduced compared to the case where time authentication is set as a condition.
- FIG. 3 is a diagram showing an example of the functional configuration of the authentication management device 10 according to this embodiment.
- the authentication management device 10 includes a speech recognition text conversion unit 101, a call reason detection unit 102, a setting classification identification unit 103, a term determination unit 104, and a time authentication setting unit 105. These units are implemented by, for example, one or more programs installed in the authentication management apparatus 10 causing a processor such as a CPU (Central Processing Unit) to execute processing.
- a processor such as a CPU (Central Processing Unit) to execute processing.
- the authentication management device 10 has a call data DB 106, a call reason DB 107, a setting condition DB 108, and a time authentication management DB 109.
- Each of these DBs is implemented by, for example, a storage device such as an auxiliary storage device.
- At least one of these DBs may be realized by, for example, a database server or the like connected to the authentication management device 10 via a communication network.
- the voice recognition text conversion unit 101 converts the voice call between the operator terminal 20 and the customer terminal 50 into text by voice recognition. At this time, the speech recognition text conversion unit 101 performs speech recognition for each speaker and converts the speech into text. As a result, the operator's voice and the customer's voice are each converted into text.
- this speech recognition is repeatedly executed while a voice call is being made between the operator terminal 20 and the customer terminal 50 (for example, it may be executed every few seconds, or if the customer or the operator speaks). It may be executed for each word or phrase.).
- the voice data representing the voice call between the operator terminal 20 and the customer terminal 50 and the text data representing the text converted from the voice call by the voice recognition text conversion unit 101 are converted, for example, after the end of the voice call.
- the call data DB 106 stores call data including voice data and text data for each call.
- the voice data includes operator voice data (hereinafter also referred to as operator voice data) and customer voice data (hereinafter also referred to as customer voice data).
- the text data includes operator text data (hereinafter also referred to as operator text data) and customer text data (hereinafter also referred to as customer text data).
- the call data also includes, for example, information such as a call ID that identifies the call.
- the call reason detection unit 102 refers to the call reason DB 107 and detects call reasons from the text converted by the speech recognition text conversion unit 101 .
- the call reason DB 107 is a database that stores information for detecting call reasons from the contents of speech of customers and operators. Details of the call reason DB 107 will be described later.
- the setting classification identifying unit 103 refers to the setting condition DB 108 and identifies the setting classification from the call reason detected by the call reason detection unit 102 .
- the setting division is a division indicating whether the entire time range from the start to the end of the call is set as the set range or a part of the time range is set as the set range.
- the setting categories are "all” indicating that the setting range is the entire temporal range from the start to the end of the call, and "" indicating that the setting range is a part of the entire temporal range shall be taken either as "partial".
- the setting condition DB 108 is a database that stores information for determining target data and setting ranges according to call reasons and important terms. Details of the setting condition DB 108 will be described later.
- the term determining unit 104 refers to the setting condition DB 108 and refers to the text converted by the speech recognition text converting unit 101 for important Determine whether the term appears.
- the time stamp setting unit 105 refers to the setting condition DB 108 and sets time stamps for all or part of the target data. That is, when the setting category specifying unit 103 specifies that the setting category is “all”, the time stamp setting unit 105 refers to the setting condition DB 108 and corresponds to the call reason detected by the call reason detecting unit 102. Set the time authentication with all the target data to be set as the setting range. On the other hand, when the setting category specifying unit 103 specifies that the setting category is “partial”, the time stamp setting unit 105 refers to the setting condition DB 108 to determine the call reason detected by the call reason detecting unit 102. and an important term appearing in the text converted by the speech recognition text conversion unit 101, and time authentication is set for the target data and the setting range.
- Time authentication settings are specifically performed as follows. That is, the time stamp setting unit 105 calculates the hash value of the setting range of the target data, and requests the time stamp token of this hash value from the time stamp authority 60 . Then, when the time stamp token is returned from the time stamp authority 60, the time stamp setting unit 105 creates time stamp information including the time stamp ID, call ID, call reason ID, and the time stamp token. Time authentication information is saved in the time authentication management DB 109 . As a result, the time authentication information associated with the call data by the call ID is saved in the time authentication management DB 109, and the time authentication is set for the setting range of the target data that is at least a part of the call data. Become. Note that the time stamp ID is an ID for identifying time stamp information. Details of the time stamp management DB 109 will be described later.
- FIG. 4 is a diagram showing an example of the call reason DB 107. As shown in FIG.
- the call reason DB 107 stores information (data or records) that associates call reason IDs, call reason names, and detection conditions.
- a call reason ID is an ID that identifies a call reason.
- the call reason name is the name of the call reason.
- the detection condition is a condition for the call reason detection unit 102 to detect call reasons.
- the information on the first line of the call reason DB 107 shown in FIG. 4 is the call reason ID "a-1", the call reason name "product guide”, and the detection condition "both: insurance”.
- the information on the second line of the call reason DB 107 shown in FIG. 4 is the call reason ID "a-2”, the call reason name "contract procedure”, and the detection condition "both: contract”.
- the detection condition is expressed in the form of "speaker: word”. This means that the detection condition is satisfied when the speaker utters the word.
- the detection condition "both: insurance” means that the detection condition is satisfied when both the operator and the customer utter the word "insurance”.
- the detection condition "customer: address change” means that the detection condition is satisfied when the customer utters the word "address change”.
- the detection condition "operator: A” means that the detection condition is satisfied when the operator utters the word "A”.
- the format of the detection condition described above is just an example, and the detection condition may be expressed in various formats (for example, a logical expression, etc.).
- FIG. 5 is a diagram showing an example of the setting condition DB 108. As shown in FIG.
- the setting condition DB 108 stores information (data or records) that associates call reason IDs, setting categories, important terms, target data, and setting ranges.
- a call reason ID is an ID that identifies a call reason.
- the setting division is a division indicating whether the entire time range from the start to the end of the call is set as the set range or a part of the time range is set as the set range. Important terms are words, phrases, etc. that are important with respect to the call reason.
- the target data is data to be set for time stamping.
- the setting range is a temporal range that is the setting range for time authentication in the target data. Important terms are not included in the information whose setting category is "all" (or NULL, null values, etc. are included as important terms).
- the information on the first line of the setting condition DB 108 shown in FIG. range "three minutes before and after important term"), (important term “guarantee", target data "text”, set range "five minutes before and after important term”), . . . ).
- the set division of the call data for the call reason ID "a-1" is "partial"
- the text data and voice data are targeted for 3 minutes before and after the speech.
- when the important term "guarantee” is uttered text data is the target data, and 5 minutes before and after the utterance is the setting range for time authentication.
- the information on the second line of the setting condition DB 108 shown in FIG. be. This is because the setting classification of the call data of the call reason ID "a-2" is "all", and the entire period (that is, the entire time range) of the text data and voice data is the setting range of time authentication. means to become
- the above setting range is just an example, and it is possible to define various temporal ranges.
- the time span before and after the utterance of the important term such as "three minutes before the utterance of the important term” and "five minutes after the utterance of the important term”.
- a definition such as "five minutes after one minute has passed since the important term was uttered” is possible.
- a clause containing an important term or a sentence break may be defined as a set range, or a range starting or ending at a conjunction that is a turning point of conversation, or the relevant conjunction as a reference.
- a predetermined range of points may be defined as the setting range, or a series of conversations between the operator and the customer, such as contractor confirmation and address confirmation, is extracted by natural language analysis, and then extracted.
- a range representing a specific scene may be defined as the setting range.
- FIG. 6 is a flowchart showing an example of the flow of time authentication processing.
- the call reason detection unit 102 refers to the call reason DB 107 and detects the call reason from the text represented by the text data included in the call data (step S101). That is, the call reason detection unit 102 refers to each record stored in the call reason DB 107 and determines whether or not there is a record whose text satisfies the detection condition. Then, the call reason detection unit 102 identifies the call reason ID and the call reason name of the record determined to satisfy the detection condition. As a result, the call reason with this call reason ID and call reason name is detected.
- the setting category specifying unit 103 refers to the setting condition DB 108 and specifies the setting category from the call reason detected in step S101 (step S102). That is, the setting classification specifying unit 103 refers to the record of the call reason ID detected in step S101 among the records stored in the setting condition DB 108, and determines whether the setting classification included in this record is "all". or "part".
- the term determination unit 104 refers to the setting condition DB 108 and determines whether or not important terms appear in the text (step S103). That is, the term determination unit 104 refers to the record of the call reason ID detected in step S101 among the records stored in the setting condition DB 108, and determines whether the important terms included in this record are included in the text. (step S103).
- the time stamp setting unit 105 refers to the setting condition DB 108, and for each important term that appears in the text, the target data corresponding to the important term and A setting range is specified (step S104). That is, the time stamp setting unit 105 refers to the record of the call reason ID detected in step S101 among the records stored in the setting condition DB 108, and for each important term appearing in the text, Identify the target data and setting range corresponding to the important term.
- time authentication setting The unit 105 specifies the target data "text, voice” and the setting range “three minutes before and after the important term” and the target data "text” and the setting range "five minutes before and after the important term”.
- step S103 the authentication management device 10 terminates the time authentication processing. In this case, time authentication is not set.
- the time stamp setting unit 105 sets time stamps for the target data and setting range corresponding to each important term appearing in the text (step S105). That is, the time stamp setting unit 105 extracts the setting range from the target data corresponding to the important term for each important term appearing in the text, calculates the hash value, and then stamps the hash value.
- a token is requested from the time stamping authority 60. Then, when the time stamp token related to the important term is returned from the time stamping authority 60, the time stamping setting unit 105 creates a time stamping ID, and then adds this time stamping ID, the call ID of the call data, Create time authentication information including the call reason ID, the important term, the target data and setting range corresponding to the important term, and the time stamp token.
- time stamp information is created for each important term that appears in the text. For example, when the important terms “fee” and “guarantee” appear in the text, time authentication information corresponding to “fee” and time authentication information corresponding to "guarantee” are created.
- the important terms corresponding to each of these multiple setting ranges may not be distinguished, or these multiple setting ranges may be combined into one.
- the time stamp setting unit 105 refers to the setting condition DB 108, and refers to the target data and setting range corresponding to the call reason detected in step S101 above. is specified (step S106). That is, the time authentication setting unit 105 refers to the record of the call reason ID detected in step S101 among the records stored in the setting condition DB 108, and determines the target data and setting range included in this record. Identify.
- the time authentication setting unit 105 specifies the target data "text, voice” and the setting range "whole period".
- the setting category is specified as "all" in step S102 above, the setting range does not necessarily have to be specified. This is because the setting category "all" means that the time stamp is set with the entire period of the target data set as the setting range.
- the time stamp setting unit 105 sets time stamps for the entire target data (step S107). That is, the time authentication setting unit 105 calculates the hash value of the setting range of the target data specified in step S106 (that is, the entire period of the target data), and converts the time stamp token of this hash value to the time A request is made to the certificate authority 60 . Then, when the time stamp token is returned from the time authentication authority 60, the time authentication setting unit 105 creates a time authentication ID, and then creates this time authentication ID, the call ID of the call data, and the call reason Create time authentication information including the ID, the target data and setting range corresponding to the call reason ID, and the time stamp token. As a result, the time authentication information for the entire target data is created.
- the time stamp setting unit 105 saves the created time stamp information in the time stamp management DB 109 (step S108).
- the time authentication is set for the set range of the target data that is at least part of the call data (including the case where the set range is the entire period).
- FIG. 7 shows an example of the time stamp information stored in the time stamp management DB 109.
- the time authentication information on the first line of the time authentication management DB 109 shown in FIG. voice”, the setting range “3 minutes before and after the important term”, and the time stamp token “...”.
- This uses the text data and voice data included in the call data with the call ID "C001" as target data, and sets a time stamp token within a setting range of 3 minutes before and after the important term "charge” is uttered as time authentication. It means that there is
- the data is "text, voice", the setting range is “whole period”, and the time stamp token is "". This means that the text data and voice data included in the call data with the call ID "C002" are the target data, and the time stamp token for the entire period is set as the time authentication.
- the authentication management device 10 is included in the call data of the voice call according to the call reason of the voice call between the operator and the customer and the important terms uttered during the call.
- a part or all of the data is set as target data, and time authentication is set for all or part of the temporal range of the target data.
- time stamps are set for appropriate ranges of appropriate data according to the importance of the call reason of the call and the importance of terms spoken during the call. Therefore, for example, the cost required for time authentication can be reduced compared to the case where time authentication is unconditionally set for all call data.
- time authentication may be performed in real time.
- the start point of the setting range is specified from the interaction between the operator and the customer, and at the end point of the setting range, the end point of the call, or the end condition being satisfied by natural language processing, the above step S105 or step S107 time authentication information may be created.
- the hash value may be calculated including not only the data within the setting range but also the current time.
- the end condition by the above-described natural language processing includes, for example, the appearance of a break between sentences or clauses, the appearance of a conjunction that serves as a turning point in conversation, and the like.
- the set range is a temporal range such as before, after, and before and after the important term starting from an utterance, but the starting point may be, for example, the execution of some action by the operator. For example, if the call reason is "contract procedure", the operator may open an electronic file or the like of the contract and read out the content of the contract. Therefore, for example, starting from an action such as "opening an electronic file of clauses", a temporal range such as before, after, and before and after that may be set as a setting range.
- the action of "opening the electronic file of the terms and conditions" above can be detected, for example, by monitoring the folder in which the electronic file of the terms and conditions is stored. Therefore, for example, this modification can be realized by including information indicating the access date and time of the electronic file in the call data.
- Other examples of such actions include scrolling the agreement file to the end, in the case of text chat, the operator attaches the agreement file and sends it to the customer, or the customer sends the agreement file as an attachment. It is also possible to open a file. In the case of a video call, the operator may share and display the agreement file on the customer's screen or display it for a certain period of time.
- step S101 of FIG. 6 the call reason is detected from the text represented by the text data. You may In this case, it is possible to implement this modification by including in the call data information indicating guidance from the IVR system and information indicating the number that the customer has entered into the customer terminal 50 in response to this guidance. .
- one call has only one call reason for the sake of simplicity, but generally one call may have a plurality of call reasons.
- a customer may call a call center for two reasons: an address change procedure and a contract plan change.
- call data is divided for each call reason, and step S102 and subsequent steps are executed for each call data after division.
- Various methods of dividing call data are possible. For example, when call reason A is detected at time T1 and call reason B is detected at time T2, time T2 - ⁇ T (where ⁇ T A possible method is to divide the call data by the time width). This is because it is considered that the same call reason is being discussed at a time slightly earlier than the time when a certain call reason is detected.
- the set range of text data or voice data or both included in call data is subject to time authentication, but data related to the set range may also be subject to time authentication.
- data related to the speech recognition function for example, speech recognition model, etc.
- Speech recognition model for example, speech recognition model, etc.
- Software or programs that have performed some kind of processing on the data, or libraries used by those software or programs may also be subject to time authentication.
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Accounting & Taxation (AREA)
- General Business, Economics & Management (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Acoustics & Sound (AREA)
- Human Computer Interaction (AREA)
- Primary Health Care (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Signal Processing (AREA)
- Human Resources & Organizations (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Telephonic Communication Services (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Facsimiles In General (AREA)
- User Interface Of Digital Computer (AREA)
Abstract
Description
まず、本実施形態に係るコンタクトセンタシステム1の全体構成について、図1を参照しながら説明する。図1は、本実施形態に係るコンタクトセンタシステム1の全体構成の一例を示す図である。
本実施形態では、必要なデータ(以下、対象データともいう。)の必要な範囲(設定範囲)に対して時刻認証が設定される。ここで、対象データは、通話データに含まれる音声データ及びテキストデータのうち、時刻認証の設定対象となるデータのことである。また、設定範囲は、対象データの時間的な範囲の中で、時刻認証の設定対象となる範囲のことである。
次に、本実施形態に係る認証管理装置10の機能構成について、図3を参照しながら説明する。図3は、本実施形態に係る認証管理装置10の機能構成の一例を示す図である。
次に、コールリーズンDB107の詳細について、図4を参照しながら説明する。図4は、コールリーズンDB107の一例を示す図である。
次に、設定条件DB108の詳細について、図5を参照しながら説明する。図5は、設定条件DB108の一例を示す図である。
以下では、オペレータ端末20と顧客端末50との間の音声通話を表す音声データと、この音声通話を音声認識テキスト変換部101により変換したテキストを表すテキストデータと、当該音声通話の通話IDとが含まれる通話データが通話データDB106に格納されているものとして、或る通話データに関して時刻認証を設定する場合について、図6を参照しながら説明する。図6は、時刻認証処理の流れの一例を示すフローチャートである。
以下、本実施形態の変形例をいくつか説明する。
本実施形態では、重要用語が発話を起点としてその前、後、前後といった時間的範囲を設定範囲としたが、例えば、オペレータの何等かの動作の実行を起点としてもよい。例えば、コールリーズンが「契約手続」である場合には、オペレータは約款の電子ファイル等を開いて、その約款内容の読み上げ等を行うことがある。そこで、例えば、「約款の電子ファイルを開く」といった動作を起点としてその前、後、前後といった時間的範囲を設定範囲としてもよい。
図6のステップS101ではテキストデータが表すテキストからコールリーズンを検出したが、例えば、IVR(Interactive Voice Response)システムからの案内に応じて顧客が顧客端末50に入力した番号からコールリーズンを特定及び検出してもよい。この場合、IVRシステムからの案内を示す情報と、この案内に応じて顧客が顧客端末50に入力した番号を示す情報とを通話データに含めることで、本変形例を実現することが可能である。
本実施形態では、簡単のため、1つの通話には1つのコールリーズンのみが存在するものとしたが、一般には、1つの通話に複数のコールリーズンが存在することもある。例えば、顧客が住所変更手続きと契約プラン変更の2つの理由でコールセンタに電話した場合等である。
本実施形態では、通話データに含まれるテキストデータ若しくは音声データ又はその両方の設定範囲を時刻認証の対象としたが、当該設定範囲に関連するデータも時刻認証の対象としてもよい。例えば、当該設定範囲内の音声データをテキストデータに変換するために使用した音声認識機能に関するデータ(例えば、音声認識モデル等)や、認証管理装置10が備えるソフトウェア又はプログラムのうち、当該設定範囲内のデータに対して何等かの処理を行ったソフトウェア又はプログラム、それらのソフトウェア又はプログラムが利用するライブラリ等も時刻認証の対象としてもよい。これにより、例えば、時刻認証が設定されたデータに対して適切な処理が行われていたのか、不具合や不都合が生じた場合の瑕疵に関する責任(不適合責任)の所在、等を検証することが可能となる。
10 認証管理装置
20 オペレータ端末
30 管理者端末
40 PBX
50 顧客端末
60 時刻認証局
70、80 通信ネットワーク
101 音声認識テキスト変換部
102 コールリーズン検出部
103 設定区分特定部
104 用語判定部
105 時刻認証設定部
106 通話データDB
107 コールリーズンDB
108 設定条件DB
109 時刻認証管理DB
Claims (12)
- 複数人の会話の内容に応じて、前記会話を表す会話データの一部又は全部を、時刻認証の設定範囲として特定する特定部と、
特定された前記設定範囲に対して時刻認証を設定する設定部と、
を有する情報処理装置。 - 前記特定部は、
第1の話者から第2の話者への問い合わせ内容に応じて、前記会話データの一部又は全部を前記設定範囲として特定する、請求項1に記載の情報処理装置。 - 前記会話の内容から前記問い合わせ内容を検出する検出部を有し、
前記特定部は、
検出された前記問い合わせ内容に応じて、前記会話データの一部又は全部を前記設定範囲として特定する、請求項2に記載の情報処理装置。 - 前記会話の内容を音声認識によりテキストに変換するテキスト変換部を有し、
前記検出部は、
前記会話の内容を変換したテキストから、前記問い合わせ内容を検出する、請求項3に記載の情報処理装置。 - 前記特定部は、
前記会話データに含まれる一部又は全部のデータのみを対象として、前記対象の時間的な範囲の全部又は一部を前記設定範囲として特定し、
前記設定部は、
前記対象の前記設定範囲に対して前記時刻認証を設定する、請求項2乃至4の何れか一項に記載の情報処理装置。 - 前記特定部は、
前記問い合わせ内容と前記会話の内容における重要事項とに応じて、前記対象と前記設定範囲とを特定する、請求項5に記載の情報処理装置。 - 前記特定部は、
前記問い合わせ内容に応じて、又は、前記問い合わせ内容と前記会話の内容における重要事項とに応じて、前記対象と前記設定範囲とを特定する、請求項6に記載の情報処理装置。 - 前記特定部は、
前記重要事項が発生した時刻を起点する所定の範囲を前記設定範囲として特定する、請求項6又は7に記載の情報処理装置。 - 前記重要事項には、予め決められた重要な用語の発話と、予め決められた重要な動作の実行との少なくとも一方が含まれる、請求項8に記載の情報処理装置。
- 前記特定部は、
前記会話の内容を変換したテキスト中で所定の用語を含む文節又は文の区切りで表される範囲、所定の接続詞を始点及び終点とする範囲、前記接続詞を基準点と所定の範囲、前記テキストから自然言語解析により抽出した範囲のうち、特定のシーンを表す範囲、の少なくとも1つを前記設定範囲として特定する、請求項4に記載の情報処理装置。 - 複数人の会話の内容に応じて、前記会話を表す会話データの一部又は全部を、時刻認証の設定範囲として特定する特定手順と、
特定された前記設定範囲に対して時刻認証を設定する設定手順と、
をコンピュータが実行する情報処理方法。 - 複数人の会話の内容に応じて、前記会話を表す会話データの一部又は全部を、時刻認証の設定範囲として特定する特定手順と、
特定された前記設定範囲に対して時刻認証を設定する設定手順と、
をコンピュータに実行させるプログラム。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2023510302A JPWO2022209144A1 (ja) | 2021-03-31 | 2022-01-14 | |
US17/761,961 US20240012994A1 (en) | 2021-03-31 | 2022-01-14 | Information processing device, information processing method and program |
GB2316355.3A GB2620348A (en) | 2021-03-31 | 2022-01-14 | Information processing device, information processing method, and program |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2021059678 | 2021-03-31 | ||
JP2021-059678 | 2021-03-31 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022209144A1 true WO2022209144A1 (ja) | 2022-10-06 |
Family
ID=83458541
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2022/001145 WO2022209144A1 (ja) | 2021-03-31 | 2022-01-14 | 情報処理装置、情報処理方法及びプログラム |
Country Status (4)
Country | Link |
---|---|
US (1) | US20240012994A1 (ja) |
JP (1) | JPWO2022209144A1 (ja) |
GB (1) | GB2620348A (ja) |
WO (1) | WO2022209144A1 (ja) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006268149A (ja) * | 2005-03-22 | 2006-10-05 | Toshiba Corp | スキャナ装置およびこのスキャナ装置を備えた文書管理システム |
JP2007081526A (ja) * | 2005-09-12 | 2007-03-29 | Matsushita Electric Ind Co Ltd | Ip電話装置、ip電話システム、及びipファクシミリ装置 |
JP2007135120A (ja) * | 2005-11-14 | 2007-05-31 | Animo:Kk | 音声証書データ生成方法及びシステム |
JP2008113165A (ja) * | 2006-10-30 | 2008-05-15 | Fuji Xerox Co Ltd | 情報処理装置、情報処理プログラム、及び情報処理システム |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11120226B1 (en) * | 2018-09-04 | 2021-09-14 | ClearCare, Inc. | Conversation facilitation system for mitigating loneliness |
US11521620B2 (en) * | 2020-02-21 | 2022-12-06 | BetterUp, Inc. | Synthesizing higher order conversation features for a multiparty conversation |
US11557288B2 (en) * | 2020-04-10 | 2023-01-17 | International Business Machines Corporation | Hindrance speech portion detection using time stamps |
US11645466B2 (en) * | 2020-07-27 | 2023-05-09 | Bytedance Inc. | Categorizing conversations for a messaging service |
US11190641B1 (en) * | 2020-09-30 | 2021-11-30 | Capital One Services, Llc | Automated agent behavior recommendations for call quality improvement |
-
2022
- 2022-01-14 GB GB2316355.3A patent/GB2620348A/en active Pending
- 2022-01-14 JP JP2023510302A patent/JPWO2022209144A1/ja active Pending
- 2022-01-14 US US17/761,961 patent/US20240012994A1/en active Pending
- 2022-01-14 WO PCT/JP2022/001145 patent/WO2022209144A1/ja active Application Filing
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006268149A (ja) * | 2005-03-22 | 2006-10-05 | Toshiba Corp | スキャナ装置およびこのスキャナ装置を備えた文書管理システム |
JP2007081526A (ja) * | 2005-09-12 | 2007-03-29 | Matsushita Electric Ind Co Ltd | Ip電話装置、ip電話システム、及びipファクシミリ装置 |
JP2007135120A (ja) * | 2005-11-14 | 2007-05-31 | Animo:Kk | 音声証書データ生成方法及びシステム |
JP2008113165A (ja) * | 2006-10-30 | 2008-05-15 | Fuji Xerox Co Ltd | 情報処理装置、情報処理プログラム、及び情報処理システム |
Also Published As
Publication number | Publication date |
---|---|
GB202316355D0 (en) | 2023-12-13 |
US20240012994A1 (en) | 2024-01-11 |
GB2620348A (en) | 2024-01-03 |
JPWO2022209144A1 (ja) | 2022-10-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11210461B2 (en) | Real-time privacy filter | |
US9009785B2 (en) | System and method for implementing adaptive security zones | |
US9571652B1 (en) | Enhanced diarization systems, media and methods of use | |
US10515638B1 (en) | System, method, and computer-readable medium that facilitate voice biometrics user authentication | |
US11862148B2 (en) | Systems and methods to analyze customer contacts | |
US10229676B2 (en) | Phrase spotting systems and methods | |
US11843719B1 (en) | Analysis of customer interaction metrics from digital voice data in a data-communication server system | |
EP2124427B1 (en) | Treatment processing of a plurality of streaming voice signals for determination of responsive action thereto | |
AU2009202016B2 (en) | System for handling a plurality of streaming voice signals for determination of responsive action thereto | |
JP2016143909A (ja) | 通話内容分析表示装置、通話内容分析表示方法、及びプログラム | |
WO2022209144A1 (ja) | 情報処理装置、情報処理方法及びプログラム | |
CN117424960A (zh) | 智能语音服务方法、装置、终端设备以及存储介质 | |
JP7274315B2 (ja) | コールセンタシステムおよび通話監視方法 | |
US10924611B2 (en) | Voice recognition system and call evaluation setting method | |
US20240259339A1 (en) | Communications apparatus using channel-communications management | |
US11418647B1 (en) | Presenting multiple customer contact channels in a browseable interface | |
CA2665014C (en) | Recognition processing of a plurality of streaming voice signals for determination of responsive action thereto | |
JP2018170611A (ja) | コールセンタシステム及び通話監視方法 | |
US9232058B1 (en) | System, method, and computer readable media for confirmation and verification of shipping address data associated with a transaction | |
WO2022209143A1 (ja) | 情報処理装置、情報処理方法及びプログラム | |
WO2023144897A1 (ja) | 検出装置、検出方法及びプログラム | |
US20240220592A1 (en) | System and method for detecting agent sharing credentials | |
KR20200061107A (ko) | 음성 인식 기반의 고객 응대 시스템 | |
WO2023062851A1 (ja) | 情報処理装置、情報処理方法、及びプログラム | |
WO2023175991A1 (ja) | 応対品質評価装置、応対品質評価方法、及びプログラム |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 17761961 Country of ref document: US |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 22779376 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 202316355 Country of ref document: GB Kind code of ref document: A Free format text: PCT FILING DATE = 20220114 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2023510302 Country of ref document: JP |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 22779376 Country of ref document: EP Kind code of ref document: A1 |